Acute Promyelocytic Leukemia (APL)

The characteristic morphology of APL blasts usually allows the prompt diagnosis of APL (examples in eLCH - eLearning Curriculum Hematology for Bone Marrow Cytology using Virtual Microscopy; https://ehaematology.com/). Microscopically, two subtypes are distinguished, the much more common hypergranular form (AML M3) and the rare hypo- (micro-) granular variant (M3v). The most important features are listed in Table 3.

If the typical morphological features of APL are present, the underlying genetic defect corresponds in about 98% to the translocation t(15;17)(q22;q12), with the resulting fusion gene PML::RARA. Also rare but possible are cryptic genetic alterations that are invisible in conventional cytogenetics but can be detected by FISH and PCR (normal karyogram but PML::RARA-positive). The reciprocal fusion gene RARA::PML is detectable in only a subset of cases. It is therefore not appropriate for routine diagnosis. In addition to PML::RARA, other concomitant mutations with as yet uncertain influence on disease progression have been described, including in the FLT3, WT1, NRAS, ARID1A, and KRAS genes [17]. The prognostic impact of additional cytogenetic aberrations is controversial.

Only very few patients have other fusion variants, where instead of the PML gene another partner gene fuses with the retinoic acid receptor alpha gene (RARA) (see chapter 6.3.1., table 6). The most common variant is the translocation t(11;17)(q23;q21), which corresponds at the molecular level to a fusion of the ZBTB16 (former name PLZF) gene and the RARA gene. More recently, single APL forms involving other retinoic acid receptors (RARB or RARG) have also been described [12].

The high rate of early mortality, especially before the start of therapy or shortly after therapy initiation, remains a major challenge in the treatment of APL. According to data from a Swedish registry study, 63% of early deaths occurred within the first week after diagnosis [10].
Prominent risk factors for early death include older age (approximately 60 years or older) and high leukocyte/blast count before initiation of therapy, and comorbidity. Elevated creatinine and male gender have also been described as risk factors [10, 18]. Data suggest that these risk factors, originally reported under ATRA and chemotherapy, are also valid for treatment with ATO [19]. Factors influencing fatal bleeding complications are described in Chapter 6.2.5.1.

During therapy with ATRA and anthracyclines (AIDA protocol of the Italian GIMEMA and Spanish PETHEMA), the combination of pre-therapeutic leukocyte and platelet count (Sanz score) proved to be a significant risk factor for the occurrence of a relapse [20]. This internationally used score (originally developed basing on ATRA and anthracycline therapy) distinguishes three risk groups (Table 5). Concerning protocols with ATRA and anthracycline therapy, the score allows a significant estimation of relapse risk and stratification of therapy intensity [20]. Other unfavorable prognostic factors detected under therapy with ATRA and anthracyclines (e.g., FLT3 length mutations, bcr3 isoform, CD56 expression, and cytogenetic additional aberrations) have not gained significance with respect to therapy stratification.

Low-risk and intermediate-risk (each with initial leukocyte count ≤10 000/µl without consideration of platelet count) are now summarized under the term standard-risk or 'non-high-risk', as the separation into low- and intermediate-risk had not proven to be significant in ATO-based therapy. Literature results make clear that the Sanz score is not applicable to protocols that include high doses of cytosine arabinoside (Ara-C) or ATO [21- 24]. Furthermore, FLT3 length mutation or cytogenetic additional aberrations had no prognostic significance in the therapy trials with ATO and ATRA [23- 26].

First-line therapy of APL is stratified by the pre-therapeutic leukocyte count (Sanz Score, Figure 1). In the absence of high risk, current standard (chemotherapy-free) therapy consists of ATRA and ATO. For high-risk APL, conventional treatment with ATRA plus anthracycline-containing chemotherapy remains the standard of care (Figure 1). Use of ATO is only possible off-label in this setting.

In medullary or extramedullary relapse of PML::RARA-positive APL, as well as in the rare case of primary resistance (persistence of a positive PCR after consolidation), long-lasting remissions and cure can be achieved by salvage therapy. In APL relapse, the achievement of a renewed molecular remission is also the critical therapeutic goal (as it is in front-line therapy). Data indicate that when salvage therapy is initiated already in molecular relapse, a reduction in early and all-cause mortality can be assumed compared with therapy in full hematologic relapse [33, 34]. In case of cytological evidence of a possible relapse (blast detection), molecular APL diagnostics should always be repeated to exclude the development of PML-RARA negative secondary AML or MDS.

An algorithm for second-line therapy is shown in Figure 2.

The choice of therapy in the first hematologic or molecular relapse of APL depends primarily on the composition of the primary therapy and the duration of the initial remission (see Figure 2). Thus, after primary treatment with ATRA and chemotherapy, ATO-based relapse therapy is recommended and vice versa. If remission is prolonged (over approximately 2 years), primary therapy may also be repeated (not evidence-based) [32].

In relapse after primary therapy with ATRA and chemotherapy, ATO in combination with ATRA is the therapy of choice due to its high antileukemic efficacy [35]. Usually, remission induction consists of a course of ATO and ATRA. Consolidation therapy analogous to primary therapy with ATO has been established with four consolidation cycles of ATO/ATRA (see chapter 6. 1. 1. 2.).

For ATO-based primary therapy, there are no evidence-based recommendations for the treatment of relapse. The focus is on conventional ATRA-plus-chemotherapy regimens, usually including higher doses of Ara-C. Longer lasting remission durations (at least 2 years) after ATO therapy are considered as possible exceptions in which ATO can be used again [32]. Also, in individual cases with shorter remission durations that cannot be exposed to the toxicity of intensive chemotherapy or, when the time to allogeneic stem cell transplantation (SCT) needs to be bridged, re-therapy with ATO may be considered. In individual cases, testing for PML mutations with arsenic resistance may be helpful in decisions to use ATO again [36].

The further therapy steps (postconsolidation therapy) are to be adapted to the individual case. For suitable patients in second molecular remission, autologous SCT is considered the therapy of choice [37- 39]. Although molecular remission is the general consensus prerequisite for autologous SZT, long-term remissions of MRD-positive cases have also been observed [40]. Also of note are reports of cases that received second-line therapy with ATO and ATRA after primary therapy with ATRA and chemotherapy and remained in stable remission without autologous SCT [7, 41]. Here, the majority (as far as information is available) were patients with longer initial remission duration [41]. These observations may be helpful in making difficult case-by-case decisions.

Pronounced risk constellations, such as the rare failure to achieve molecular remission after first-line therapy or second and subsequent relapses, usually represent an indication for the use of allogeneic SCT [39]. If allogeneic SCT is not feasible, the decision must be made on an individually. In this case, further treatment with ATO or chemotherapy or with the combination of both, if necessary as 'low dose' continuous therapy or experimental therapy can be performed (Figure 2). In this context, special reference should be made to the high efficacy of the toxin-coupled anti-CD33 antibody gemtuzumab ozogamicin (GO) in multiply pretreated APL [42]. GO is approved by the U.S. Food and Drug Administration (FDA) and subsequently by the EMA for selected AML patients, but not for APL. Starting in Japan, synthetic retinoids (tamibarotene, not available in Germany) with improved efficacy compared to ATRA were used [43].

The CNS is by far the most common site of extramedullary relapse. These are isolated CNS relapses or hematologic or molecular bone marrow relapses involving the CNS. Although there are no formal data on the use of intrathecal (i.th.) CNS prophylaxis (either for chemotherapy or ATO-based therapy), if chosen, it should be restricted to high-risk APL. In particular, if the leukocyte count is high, CNS prophylaxis should be given only after hematologic remission has been achieved to avoid complications, especially bleeding [3, 32].

For local CNS relapse, repeated weekly i.th. triple therapy with methotrexate, Ara-C, and hydrocortisone is recommended until blast clearance, followed by 6 to 10 dilated applications for consolidation [3]. For prophylaxis or treatment of the usually subsequent or simultaneous medullary relapse, systemic therapy with good CNS penetration (ATO or high-dose ARA-C) should be administered. For consolidation with autologous or allogeneic SCT, the recommendations of medullary relapse apply. In addition, there is the option of craniospinal irradiation [3, 32].

Molecular monitoring by quantitative REAL-time PCR (qPCR) of PML::RARA is an obligatory control parameter to assess treatment response as well as for early detection of molecular relapse.

The primary goal of molecular monitoring is to record the remission status after the end of consolidation therapy and, in the further course, the early detection of a molecular relapse even before the transition to a hematological relapse.
For standard risk APL, the following approach is recommended depending on the therapy:

1. in case of therapy with ATRA and ATO, a routine control by means of bone marrow aspirate with the aim of detecting molecular remission is obligatory after the end of consolidation therapy. (The individual course of the values of qPCR during the induction and consolidation phase has no influence on the therapy management, the result after consolidation therapy is essential). Once molecular remission is achieved after consolidation, follow-up monitoring is no longer routinely recommended due to the very low recurrence rate (<5%).

2. in therapy with ATRA and chemotherapy, the clinical benefit of molecular monitoring is assessed differently, so that the decision here must be made individually [32, 44].

In high-risk APL, molecular follow-up should always be performed after molecular remission has been achieved, regardless of the primary therapy chosen. It is also recommended after relapse [32, 44].

Molecular monitoring can be performed either by 3-monthly controls from bone marrow aspirate or by shorter-term controls (e.g., 4- to 6-weekly) from peripheral blood, although analysis from peripheral blood is less sensitive. There is no evidence for the duration of monitoring. Given the relative rarity of subsequent relapses, a duration of 2 years is considered sufficient [32, 44].

With qPCR, which is the current standard method for molecular monitoring, the quantity of transcripts and the kinetics of MRD is detectable much more accurately than with nested RT-PCR, which was used formerly. (Should nested RT-PCR still be used, a sensitivity of at least 10^-4 is required.)

In case of recurrence of a positive transcript detection after originally achieved qPCR negativity or an increase above 1Log, a developing relapse is suspected. In such cases, short-term monitoring of the results (after approximately 2 to 4 weeks) in the context of a new bone marrow aspiration is recommended. However, interpretation of the result in high-sensitivity qPCR can sometimes be difficult if the number of transcripts is low. In this case, the increase in transcripts with further controls is the most reliable marker of "true" positivity [32].

ATRA, a derivative of vitamin A acid, causes an abrogation of the differentiation block of promyelocytic blasts at the molecular level and induces maturation into mature neutrophilic granulocytes, associated with a regression of coagulopathy within a few days. The exceptionally high efficacy of ATRA is APL-specific. ATRA is an obligatory component of induction therapy in APL. With ATRA monotherapy, 80% to 90% of patients with newly diagnosed APL achieve hematologic remission. Because ATRA alone does not induce durable remissions, ATRA monotherapy is usually restricted to palliative treatment goals.

Therapeutic intervention under ATRA monotherapy is primarily required in the occurrence of the APL differentiation syndrome (ADS, formerly ATRA syndrome) and in the event of a rapid leukocyte increase (Chapter 6.2.5.2. and Chapter 6.2.5.3.). Pseudotumor cerebri is observed mainly in children and young adults and is a potentially threatening complication requiring dose reduction of ATRA [3]. Dose adjustments may also be required in the presence of impaired hepatic or renal function.

Arsenic compounds represent the most effective monosubstance with curative potential in APL. The greatest clinical experience is with arsenic trioxide (As₂ O₃ ; ATO). ATO has predominantly apoptosis-inducing properties in addition to a differentiating effect. ATO targets the PML portion of the PML::RARA protein, whereas the RARA portion is the target of ATRA. The combination of the two synergistically acting compounds enhances antileukemic efficiency and leads to degradation of the PML::RARA fusion protein and, as a consequence, to apoptosis of APL blasts. Complete eradication of the APL clone (curative effect) can only be achieved by ATO [3, 11].

The onset of action of ATO is accompanied in about 50% of cases by an increase in the leukocyte count (hyperleukocytosis, see chapter 6.2.5.2.) and by increased numbers of precursors of granulopoiesis in the peripheral blood. A potentially dangerous side effect of therapy is ADS, which is observed in up to 25% of cases (see Chapter 6.2.5.3.). Concurrent use of ATRA may increase the risk of ADS. During therapy with ATO, ECG changes with prolongations of QTcF time (Fridericia correction method, Table 2) and electrolyte shifts requiring substitution, especially of potassium and magnesium, require special attention (see Chapter 5.1). Potassium should be above 4 mmol/l and magnesium above 1.8 mg/dl. Regular ECG checks are indicated. If the QTcF interval exceeds 500 msec, therapy is discontinued because of the risk of cardiac arrhythmias (torsade des pointes) [32]. After regression, treatment should be continued at 50% of the previously given daily dose (for further information on ECG and electrolyte monitoring, see EMA approval text of ATO: Trisenox, INN-arsenic trioxide (europa.eu)).

Comedication with drugs that, like ATO, can prolong QTcF time should be avoided and, if unavoidable, requires intensified ECG monitoring.

Relatively common, usually non-life-threatening side effects include liver dysfunction and increase in transaminases, nausea, vomiting, exanthema, fatigue, fever, neuropathy, and diarrhea. Reactivations of herpes zoster infections are more frequent. Regarding herpes prophylaxis (including vaccination), there are no firm recommendations. Detailed information and instructions for action in case of adverse effects of ATO are provided in the expert information of arsenic trioxide (Trisenox^TM) [45].

Oral arsenic compounds are highly effective, as is the intravenous route of administration. They are not available in the EU but are widely used in China [43].

APL blasts are highly sensitive to anthracyclines. Daunorubicin, which was originally used, has now been largely replaced by idarubicin, as this agent was found to be more effective in APL therapy. Data suggest that the inclusion of medium-high to high-dose Ara-C in high-risk APL reduces the relapse rate [21, 22][25- 28]. The spectrum of side effects of chemotherapy regimens is similar to that of AML therapy.

Gemtuzumab ozogamicin (GO) is an antibody-drug conjugate. It consists of a monoclonal antibody directed against CD33 covalently bound to the cytotoxic agent calicheamicin. Calicheamicin induces double-strand breaks in target cells, comparable to the effect of conventional chemotherapy. GO was shown to be highly effective in the relapse and primary therapy of APL [42]. However, the compound can only be used off-label in APL because the approval for CD33-positive AML does not include APL.

A selection of overall very rare molecular variants of APL including RARA is shown in Table 6. The most common variant with the translocation t(11;17)(q23;q21) is usually not sensitive to therapy with ATRA or ATO. Other variants show partial sensitivity to ATRA and/or ATO. However, this experience is based only on small case collections or individual cases, which should be considered when making treatment decisions. In cases of resistance, chemotherapy is recommended as in AML. In some of these rare variants of APL, molecular monitoring by the standard method, qPCR, is not possible. In such cases, an NGS-based MRD method may be helpful to provide insight into the status of residual disease.

More recently, APL forms involving other retinoic acid receptors (RARB or RARG) have also been described. Where studied, they showed no sensitivity to ATRA or ATO [12]. A comprehensive review of genetic alterations in atypical APL was published by an Italian research group [49].

Even in advanced age, a curative therapy approach should be pursued in the case of APL [9, 50]. The sensitivity of blasts to therapy does not differ from that of younger patients. Comorbidity and therapy-associated toxicity, especially when conventional ATRA-plus-chemotherapy concepts are used, are therapy-limiting in older patients. According to an analysis of registry data, older patients benefit from the use of ATO-based therapy in terms of remission and relapse rates. This applies to cases with standard and high-risk APL (off-label for high-risk APL) [9].

Secondary, therapy-associated APL should be treated as primary APL with the goal of cure. Prior anthracycline exposure is to be considered in the choice of therapy. The use of ATO-based primary therapy proved to be advantageous over ATRA-plus chemotherapy in a retrospective comparison, allowing a curative approach without administration of further chemotherapy [9].

The care of pregnant APL patients must be interdisciplinary. Even if APL is diagnosed during pregnancy, the patient has a chance of recovery. The stage of pregnancy is decisive for the therapeutic procedure. While it is usually not possible to successfully terminate the pregnancy in the first trimester, there is a good chance to successfully terminate the pregnancy in the second and especially in the last trimester.

ATRA and ATO have a high teratogenic potential. Options in the first trimester are termination of pregnancy (cave bleeding complications) or mono-chemotherapy with daunorubicin. After abortion, standard risk-adapted therapy can be started immediately.

In the second and third trimesters, there are no contraindications to combined treatment with ATRA and anthracyclines. A summary of published cases of all AML types shows no increased maternal risk and no increased risk of malformations in the child. However, there is an increased rate of miscarriage, premature birth, and low birth weight neonates [51, 52]. Because these complications are associated with chemotherapy, in low- or intermediate-risk pregnant women with APL, the time until after delivery can be bridged by monotherapy with ATRA, if possible. In high risk patients, combination therapy of ATRA and anthracyclines (preferably daunorubicin) is indicated despite the associated risks. In all cases, close follow-up and gynecologic co-management are required.